US6427172B1 - Method and apparatus for internet cache content delivery via a data distribution system - Google Patents

Abstract

A cache system and caching methods that are used to improve data delivery service provided by a data distribution system. The present invention uses a master cache in which predictive harvesting of content is performed based upon probability distributions of individual caches satisfying requests from users. Content in the master cache is distributed to user sites by way of the data distribution system, and local caches at user sites are automatically loaded with incoming content.

Description

BACKGROUND

The present invention relates generally to data distribution systems, and more particularly, to improved methods and apparatus for providing Internet cache content delivery using a data distribution system.

The assignee of the present invention has developed a data distribution system, and in particular, a satellite-based data distribution system, that is used to distribute data supplied by content providers to personal computers of clients. In order to allow clients to have improved response times to obtain data distributed by the data distribution system, an important aspect of the system was developed relating to caching services that addresses this need.

Several companies, including Inktomi, Cache Flow, and NetWork Appliance, for example, currently make cache appliances or specific software which are targeted at Internet service providers. SkyCache is currently operating caching services using satellite communications.

It would be an improvement to have a method and apparatus that provides Internet cache content delivery using a data distribution system.

SUMMARY OF THE INVENTION

The present invention provides for a cache system and caching method that is used to improve data delivery service provided by a data distribution system, such as a satellite-based data distribution system. The present invention uses a master cache in which predictive harvesting of content is performed based upon probability distributions of individual caches satisfying requests from their users (clients). Content in the master cache is distributed to user sites by way of the satellite-based data distribution system, and local caches at user sites are automatically loaded with incoming content. The use of the satellite-based data distribution system provides an efficient economic multiplier.

More specifically, an exemplary system comprises a master cache that receives content for distribution by the data distribution system to one or more users. A gateway receives content distributed by the data distribution system from the master cache. One or more local caches store the content received by the gateway destined for the one or more users. Harvesting software processes information from the master cache and the gateway that corresponds to probability distributions that the local caches satisfy requests from their respective users to predictively distribute the desired content to the respective users.

Content distributed by the data distribution system to the gateway is transferred into the local cache by creating a pseudo client on the gateway, receiving an interrupt at the pseudo client indicating that that content has arrived at the gateway, enabling the gateway as a sibling cache for the local cache, requesting content to be transferred from the sibling cache to the local cache, verifying that content has been transferred to the local cache during the transfer process, and disabling the gateway as a sibling cache of the local cache at the completion of the process. The local cache retrieved the content from the sibling cache until all content has been transferred.

The master cache is built by processing statistics derived from the master cache and the local cache to produces a list of content to add to the master cache and a list of content to delete from the master cache. A pseudo client is formed that retrieves and verifies the content to be added to the master cache. The verified content is then transmitted from the master cache to the local cache.

Caching is used to improve the response time of a user (client) that is served by the cache system employed in the data distribution system. Caching also benefits service providers by reducing communications costs between the service environment and the external world. Using satellite transmission services allows cost savings and response times to be further enhanced for a large number of service providers by allowing aggregation of their individual requests.

The use of the satellite-based data distribution system provides for efficient distribution of cache content over a wide geographical area thereby allows many user sites to simultaneously update their local caches. This reduces cost of terrestrial infrastructure for the individual user sites.

BRIEF DESCRIPTION OF THE DRAWINGS

The various features and advantages of the present invention may be more readily understood with reference to the following detailed description taken in conjunction with the accompanying drawings, wherein like reference numerals designate like structural elements, and in which:

FIG. 1 illustrates the architecture of an exemplary caching system in accordance with the principles of the present invention used in conjunction with a satellite-based data distribution system;

FIG. 2 illustrates exemplary apparatus and a method in accordance with the present invention for forcing content into a local cache; and

FIG. 3 illustrates an exemplary method in accordance with the present invention for building the master cache.

DETAILED DESCRIPTION

Referring to the drawing figures, FIG. 1 illustrates the architecture of anexemplary caching system10 in accordance with the principles of the present invention the invention is used in conjunction with adata distribution system30, such as a satellite-baseddata distribution system30 developed by the assignee of the present invention or an RF data distribution system, for example. An exemplary satellite-baseddata distribution system30 comprises anetwork operations center11 that includes subsystems and software that permits use of thedata distribution system30 by subscribers (users). This is illustrated assubscriber entitlement12.

The subscriber entitlement subsystem originates entitlement messages, which activates the entitled gateways to decode the data transmitted over thedata distribution system10.

The satellite-baseddata distribution system30 distributes content received from service providers by means of asatellite multicast feed13 distributed by way of asatellite14 to agateway15, referred to as a CyberStar Enterprise Gateway15, and which is part of thepresent caching system10. The content is typically transmitted in the form of http (hypertext transport protocol) objects such as web pages or nttp (News text transport protocol) objects such as news groups.

Content received by thegateway15 is transferred to alocal cache20 which is typically located on a user computer or server computer at a user site. For example, the content is stored in thelocal cache20 on a server computer and is transferred to a user computer that is coupled to auser network24 by way of aswitch22, for example. Arouter23 is used to connect theuser network24 to the Internet25.

Theexemplary caching system10 also includes amaster cache18 andharvesting software17 that is typically disposed on acomputer19 at thenetwork operations center11. Thecomputer19 is coupled to the Internet25 and receives content from content providers for distribution. Thecomputer19 is coupled by way of abackchannel16 to thegateway15. Thebackchannel16 may be implemented in a variety of ways, including a dedicated narrowband circuit (telephone), a virtual private narrowband channel, or an internet connection, for example. Thebackchannel16 is used to transmit hit/miss data and probability tables from thegateway15 to thecomputer19, and theharvesting software17 processes information contained in the transmit hit/miss data and probability tables.

FIG. 2 illustratesexemplary apparatus40 and amethod50 in accordance with the present invention for forcing content into thelocal cache20. Theapparatus40 comprises thegateway15 which is coupled to thelocal cache20 as discussed with reference to FIG.1. Themethod50 is implemented in thegateway15 and processes content received from thesatellite14, typically in the form of a multicast satellite feed. Themethod50 comprises the following steps.

Receivedcontent51 is forced into thelocal cache20 using themethod50 as follows. An entitlement message is received52 by thegateway15 from the subscriber entitlement subsystem that notifies53 apseudo client41 that new objects have arrived. Thepseudo client41 is a computer program which runs on the gateway computer and uses the appropriate communication protocol to communicate with the local cache. An example of this protocaol is ICP (Internet Cache Protocol). However, the protocol may also be a proprietary one from a number of cache hardware vendors. Thepseudo client41 also has processes that allow it to control or designate the presedence relationship between the local cache and the gateway.

Thepseudo client41 enables54 thegateway15 as asibling cache42 for thelocal cache20. The standard Internet Protocol (IP), or Internet Cache Protocol (ICP), is used to communicate between thelocal cache20 and thesibling cache42. Thepseudo client41 then requests55athenew objects53 having arrived on thesibling cache42. Thisrequest55amakes use of standard control settings which force thelocal cache20 to render any objects resident at the local cache as unable to satisfy therequest55a. Thelocal cache20 then requests55bthe object from itssibling cache42. Thesibling cache42 complies by sending56 the object to thelocal cache20. Thelocal cache20 in turn forwards the object to thepseudo client41 thus verifying57 the object has been transferred to thelocal cache20. Thepseudo client41 continues to request55aobjects from thelocal cache20 until all the new objects have been transferred to thelocal cache20. During the transfer process, thepseudo client41 verifies57 that all pages have been transferred to thelocal cache20. At the completion of the process, thepseudo client41 disables58 thegateway15 as asibling cache42 of thelocal cache20.

FIG. 3 illustrates an exemplary method60 in accordance with the present invention for building themaster cache18. Theharvest analysis software17processes statistics61 from themaster cache18 and alllocal caches20. Parameters and statistics used by the harvest analysis software are, for example, time to live (TTL) the number of times an object was requested, the number of times an object was requested and satisfied by local cache20 (Hit), the number of times an object was requested and satisfied by a server other than the local cache20 (miss). Objects are added/dropped according to high miscounts and low hit counts. The TTL parameter can also be adjusted to reduce the object's miss count. The statistics are collected by thelocal cache20 as part of its normal operation. These statistics are periodically transferred to thegateway15.

Theharvest analysis software17 produces alist62 of objects (corresponding to each item of content) to add to themaster cache18 and alist64 of objects to delete from themaster cache18. Apseudo client41ais used to retrieve65 and verify66 the objects to be added to the master cache. The master cache forwards67 the content to the satellite NOC for transmission according to various distribution policies. The satellite NOC adds a level of conditional access onto the transmission stream to provide service to only those sites which have been entitled to the cache service.

Thus, an improved method and apparatus for providing Internet cache content delivery using satellite transmission services have been disclosed. It is to be understood that the described embodiments are merely illustrative of some of the many specific embodiments that represent applications of the principles of the present invention. Clearly, numerous and other arrangements can be readily devised by those skilled in the art without departing from the scope of the invention.

Claims (19)

What is claimed is:

1. A caching system for use with a data distribution system, comprising:

a master cache for receiving content for distribution by the data distribution system to one or more users;

a gateway for receiving content that is distributed by the data distribution system from the master cache;

one or more local caches for storing the content received by the gateway destined for the one or more users; and

harvesting software coupled to the master cache and the gateway for processing information corresponding to probability distributions that the local caches satisfy requests from their respective users to predictively distribute the desired content to the respective users.

2. The system recited inclaim 1 wherein the harvesting software processes information contained in transmit hit/miss data and probability tables generated at the gateway.

3. The system recited inclaim 1 wherein the content comprises http objects.

4. The system recited inclaim 1 wherein the content comprises nttp objects.

5. The system recited inclaim 1 wherein the gateway comprises:

at pseudo client for receiving an entitlement message indicating that that content has arrived at the gateway, for enabling the gateway as a sibling cache for the local cache for requesting content to be transferred from the sibling cache to the local cache, for verifying that content has been transferred to the local cache during the transfer process, for disabling the gateway as a sibling cache of the local cache at the completion of the process:

and wherein the local cache retrieves the content from the sibling cache until all content has been transferred.

6. The system recited inclaim 5 wherein the Internet Protocol is used to communicate between the local cache and the sibling cache.

7. The system recited inclaim 5 wherein the Internet Cache Protocol is used to communicate between the local cache and the sibling cache.

8. The system recited inclaim 1 wherein the harvesting software:

processes statistics derived from the master cache and the local caches to produces a list of content to add to the master cache and a list of content to delete from the master cache;

forms a pseudo client to retrieve and verify the content to be added to the master cache; and

transmits the verified content from the master cache to the local caches.

9. The system recited inclaim 1 wherein the harvesting software:

processes statistics derived from the master cache and the local caches to produces a list of content to add to the master cache and a list of content to delete from the master cache;

forms a pseudo client to retrieve and verify the content to be added to the master cache; and

transmits the verified content from the master cache to the local caches.

10. The system recited inclaim 1 wherein the data distribution system, comprises a satellite-based data distribution system.

11. A method for transferring content distributed by a data distribution system to a gateway into a local cache, comprising the steps of:

creating a pseudo client on the gateway;

receiving an interrupt at the pseudo client indicating that that content has arrived at the gateway;

enabling the gateway as a sibling cache for the local cache;

requesting content to be transferred from the sibling cache to the local cache;

verifying that content has been transferred to the local cache during the transfer process;

disabling the gateway as a sibling cache of the local cache at the completion of the process; and

causing the local cache to retrieve the content from the sibling cache until all content has been transferred.

12. The method recited inclaim 11 wherein the data distribution system, comprises ma satellite-based data distribution system.

13. A method for building a master cache used to transfer content by way of a data distribution system to a local cache, comprising the steps of:

processing statistics derived from the master cache and the local cache to produces a list of content to add to the master cache and a list of content to delete from the master cache;

forming a pseudo client to retrieve and verify the content to be added to the master cache; and

transmitting the verified content from the master cache to the local cache.

14. The method recited inclaim 13 wherein the step of processing statistics comprises the step of processing information corresponding to probability distributions that the local caches satisfy requests from their respective users to predictively distribute the desired content to the respective users.

15. The method recited inclaim 13 wherein the step of processing statistics comprises the step of processing information contained in transmit hit/miss data and probability tables.

16. A caching system for use with a data distribution system, comprising:

a master cache for receiving content for distribution by the data distribution system to one or more user computers,

a gateway that is distinct from the one or more user computers for receiving content that is distributed by the data distribution system from the master cache;

one or more local caches for storing the content received by the gateway destined for the one or more user computers; and

harvesting software coupled to the master cache and the gateway for processing information corresponding to probability distributions that the local caches satisfy requests from their respective users to predictively distribute the desired content to the respective user computers.

17. The system recited inclaim 16 wherein the harvesting software processes information contained in transmit hit/miss data and probability tables generated at the gateway.

18. The system recited inclaim 16 wherein the gateway comprises:

a pseudo client for receiving an entitlement message indicating that that content has arrived at the gateway, for enabling the gateway as a sibling cache for the local cache, for requesting content to be transferred from the sibling cache to the local cache, for verifying that content has been transferred to the local cache during the transfer process, for disabling the gateway as a sibling cache of the local cache at the completion of the process;

and wherein the local cache retrieves the content from the sibling cache until all content has been transferred.

19. The system recited inclaim 16 wherein the data distribution system, comprises a satellite-based data distribution system.

Images